Re: Triggeriness ...

>It occurs to me that it would be possible for a distinction to exist
within
>trigger languages, between ones where "Trigger role 1" = "subject",
>"Trigger role 2" = "object", and those where "Trigger role 1" =
>"Intransitive subject or transitive object", "Trigger role 2"
= "Transitive
>subject". So, while it would not be useful to classify trigger languages
as
>Nom/Acc or Erg/Abs as a primary classification, there may be scope for a
>secondary classification as "Nom/Acc-like trigger language" or
>"Erg/Abs-like trigger language".
You mean, if in trigger languages there was a
favoured voice that set a default role to the
trigger as agent or as object? As far as I know,
in Tagalog there's no favoured assignation
of role to the trigger (which in itself is
semantically 'meaningless', since by itself
doesn't correlate to any semantic role neither
in absolute terms nor by default), so you
cannot establish a parallel subdivision to
the hierarchical organization of core cases in
languages where verbs accept valencies higher
than 1. Note that both English and Basque
feature plurivalent verbs while Tagalog
doesn't, and it is in the hierarchization of
the cases in plurivalent verbal schemes where
the difference between the accusative and
ergative types arises, since in the monovalent
intransitive scheme (which is the only one
used in Tagalog) the meaning of the single
'subject' core case has, by necessity, to be
determined by the semantics of the verb.
E.g. in English the subject of intransitives
expresses different semantic roles according
to which verb it is the subject of (and sometimes
according to the animacy level of the subject
too): "I fall", "I read", "It reads (as ...)".

Though, of course, you can invent trigger-type
(by this I mean that verbs are all monovalent)
conlangs where there actually is a clearly
favoured verbal voice (e.g. because it is left
unmarked) and thus the trigger can be said to
have a default assignation of semantic role.
And then subclassify the trigger-type of those
conlangs in parallel with the accusative
and ergative types, according to which default
voice, and from it the role of the subject-trigger,
is used. Or, instead, use a classification
that considers the parameters of core-case
hierarchization and valency-ranking separately,
so that an accusative-like trigger-type would
be labelled, say, "accusative intransitive type",
the usual accusative-type would be an "accusative
transitive type" and Tagalog would be a
"non-hierarchical intransitive type".

Cheers,
Javier

Reply